Insert the title of your presentation here
Presented by Name HereJob Title - Date
Overview of LDW/AEBS research for the EC
Presented by Iain Knight9th December 2008
Informal document No. GRRF-S08-11Special GRRF brainstorming session 9 December 2008Agenda item 2(b)
Introduction
Definitions
Objectives and limitations of the studies
AEBS- System functions
- Technical requirements
- Assessing the benefits
LDW- Technical requirements
- Costs and benefits
Page 3
Definitions used in the research Lane Departure Warning (LDW) systems monitor the position of the
vehicle with respect to the lane boundary. When the vehicle is in danger of leaving the lane unintentionally, the system delivers a warning to the driver
Lane Change Assist (LCA) monitors the areas to the side and rear of the subject vehicle and warn the driver if a change of lane is commenced that could cause a collision with a vehicle in the blind spot
Lane Keeping Assistance (LKA) is a LDW that takes additional action (e.g. active steering, braking corrections) to help the driver avoid leaving the lane unintentionally
Automated Emergency Braking System (AEBS) is a generic name for any system that can apply emergency braking independent of driver control
Collision Mitigation Braking System (CMBS) is a system that can autonomously apply emergency braking in order to mitigate the severity of a collision that has become unavoidable
Collision Avoidance Braking System (CABS) is a system that can autonomously apply emergency braking in order to fully avoid a collision.
Objectives of the studies
Page 4
To gather and evaluate information regarding the technical requirements, costs and benefits of the systems, with respect to application to different vehicle types:- Light vehicles (M1 and N1);
- Heavy goods vehicles (N2 and N3)
- Large passenger vehicles (M2 and M3)
- Considering the benefits to:- Occupants of the equipped vehicle;
- Occupants of vehicles in collision with the equipped vehicle; and
- Vulnerable road users (VRU) i.e pedestrians, pedal cyclists and motorcyclists
Both studies were desk-based, limited to analysis of existing literature, consultation with industry and accident data analysis
Page 5
Key characteristics of LDW systems
Sensor technology
Should there be specific requirements for the types of sensor that can be used?
System behaviour
What speed should the system function at? What road curvature should the system function on? Where should the warning threshold be?
System capability
What type of boundaries are detectable What Weather/environmental conditions should the system function in?
Human-machine interface
How should the warning be presented? What status information should be indicated to the driver and how? How much driver control and adjustment of the system should be
permitted?
What requirements are needed in the following areas?
Page 6
Existing technical requirements (LDW)
ISO 17361:2007
Specifications, requirements and test methods for passenger cars, commercial vehicles and buses
Functional elements:
Lateral position detection
Warning
Status indication
Suppression request
Vehicle speed detection
Driver preference
FMCSA-MCRR-05-005
The Federal Motor Carrier Safety Administration Concept of Operations and Voluntary Operational Requirements (USA)
Large trucks >10,000lbs
Main functional elements same as ISO 17361 (different terminology)
Two technical standards for LDW identified
Questions for consideration (LDW)
Page 7
ISO 17361:2007 has different performance limits for commercial vehicle and cars, is this appropriate?
Current performance specifications do not include function in adverse weather conditions. Is this necessary/feasible?
Two classes of LDW are permitted, based on minimum radii and speed for which they are functional. Should both be permitted?
Warning can occur before or after lane boundary crossed. Effectiveness vs false alarm balance? Where should the regulation draw the line?
Lane boundaries in tests must be “in good conditions and in accordance with applicable national standards for lane marking design and materials” i.e. one type in good condition per country. How should this be assessed given a single approval for multiple regions and possible diversity within a region?
Page 8
Relevant accidents (LDW)
Accidents on single carriageway roads where the VOI has drifted out of the lane of travel into an oncoming lane, where a collision has occurred.
Head-on (A)
Accidents where the VOI leaves the lane in which they are travelling, resulting in the vehicle leaving the road or colliding with roadside barriers.
These accidents tend to be single vehicle, but can also involve VRU
Leaving road (B)
Accidents on carriageways with multiple lanes in the same direction. The VOI leaves the lane and there is a collision between the VOI and a vehicle in the adjacent lane (either side to side or front to rear of VOI).
Side-swipe (C)
Three groups of accidents identified
Target population data for GB and Germany extrapolated to EU
Effectiveness data taken from literature and applied to target population
Variation in GB/Germany data combined with wide range of effectiveness in literature led to wide range of predicted effects
Page 9
Estimating benefits (LDW)
Target Population (A+B+C)
Effectiveness (% of target population)
Total BenefitCasualty severity A B C
Fatal 23-181 16-48 48 16-48 4-87
Serious 157-1143 12-36 36 12-36 19-468Slight 597-2148 7-20 20 7-20 42-490
Annual casualty benefit – LDW on N2/N3 vehicles
Target Population (A+B+C)
Effectiveness (% of target population)
Total BenefitCasualty severity A B C
Fatal 7-201 16-48 48 16-48 1-96
Serious 51-1066 12-36 36 12-36 6-408Slight 373-1105 7-20 20 7-20 26-255
Annual casualty benefit – LDW on M2/M3 vehicles
Casualty valuations Fatal €1,000,000Serious €135,000Slight €15,000
Page 10
Costs
Retail pricesUnit cost used in analysis
€384 - €448 from various manufacturers information
€200-€448€300 Abele et al (2005) for 2010
€200 Abele et al. (2005) for 2020
€200 COWI used €400 for combined system
Only retail costs identified
Benefit-cost ratios (BCR)
Vehicle type Limit BCR
N2/N3 Min 0.18
N2/N3 Max 6.56
M2/M3 Min 0.47
M2/M3 Max 23.97
Assuming mandatory fitment in 2013
Characteristics of AEBS
Current systems (2006)- Mitigation systems
- Front to rear shunt collisions with other vehicles and some fixed objects- No operation at very low or very high speeds/relative speeds- Limited function in adverse weather conditions- Curve function limited to line of sight- Varying strategies – partial braking applied early to full braking applied late
- Avoidance systems- Low speed function (<20 km/h) only- Other characteristics as for mitigation systems
Future systems- Expanded functionality e.g.
- Pedestrian, junction & head on collisions (latter two may require V-V/V-I communication
Technical requirements
In 2006, only one set of Technical requirements in existence (MLIT guidelines – Japan)- Prescribed activation thresholds based on TTC, steering and braking
capability
- Defined minimum levels of automated braking
- Not all EU models would have complied
- Good basis but further development required
ISO standard under development but not available for review
No published data identified to assess whether a risk of sensor interference in situations where multiple equipped vehicles were present
Assessing the benefits - CMBS
Two extreme sets of 1st generation CMBS characteristics were defined- Partial braking applied late- Full braking applied early- Neither system expected on market but all realistic systems will fall between
the two.
UK in-depth fatal accident data analysed to predict potential effect of the two “extreme” systems fitted to HVs.- Total number of fatal accidents on database >1,800- 70 cases met selection criteria (e.g. front of HV to rear of other vehicle, not
snowing, speed information present etc.)- Collision speeds re-calculated according to system characteristics- Estimated 25%-75% of fatalities in front to rear shunts could be mitigated
Similar approach undertaken for light vehicles but insufficient cases on in-depth database for conclusive result.
Scoping the potential future benefits - AEBS
“what if” scoping study undertaken to assess the future potential of more developed systems.- Based on target population data from GB STATS19 extrapolated to EU
using EuroSTAT. Divided by- Vehicle class fitted to (M1,2,3; N1,2,3; L)
- Accident configuration- Front to rear of other vehicle
- Head on collisions
- Collisions with fixed objects on/off the carriageway
- Collisions with pedestrians
- Front to side collisions
- Casualty estimates reflect the potential IF systems could be as effective as 1st generation (HV) systems when fitted to other vehicles and when involved in different collision types (i.e. 25%-75%)
Benefits and break-even costsSystem class Vehicle class AEBS fitted to.
Current Near future Longer term
Fatality reduction 313 – 1,149 2,043 – 7,489 1,349 – 4,946 M1
Break even cost (€) 26 – 216 136 – 966 96 – 703
Fatality reduction 4 – 14 96 – 351 55 – 202 M2/3
Break even cost (€) 197 – 1,731 1,732 – 12,324 871 – 6,217
Fatality reduction 44 – 160 148 – 543 185 – 681 N1
Break even cost (€) 26 – 182 68 – 443 76 – 500
Fatality reduction 102 – 372 180 - 659 319 – 1,170 N2/3
Break even cost (€) 314 – 1,475 432 – 1,938 773 – 3,481
Fatality reduction 618 – 2,265 L
Break even cost (€)
1,322 – 5,704
Positive BCR more likely for heavy vehicles- Front to rear shunt accidents much more severe with HVs than with light vehicles
- Costs applied to c.1/50th of the number of vehicles
Conclusions
For both LDW and AEBS casualty benefits greater if fitted to cars but BCRs greater when fitted to heavy vehicles
Considerable diversity in technical specifications and performance
Particularly for AEBS, future developments have more casualty reduction potential than 1st generation if they can be developed effectively
Technical requirements are more developed for LDW than for AEBS but further development likely to be needed for both
Examples of areas for further consideration
Generations- Both concepts are likely to be developed in different “generations”
- Varying performance capabilities already exist (e.g. different classes in LDW ISO, LKA, mitigation or low speed avoidance for AEBS)
- What functions/generations should be considered in scope?
- What should the performance limits for those functions be?
- How can the requirements best be implemented without stifling future development of the next generation?
In service performance